In the rainy season of 1989, IgG and IgM antibodies against dengue and Japanese encephalitis viruses (measured by enzyme-linked immunoassay [ELISA]) in serum from all primary-school children in two areas of central Thailand were sampled in order to choose a study site for more detailed epidemiological and entomological analysis. Students in three schools in the largely non-agricultural, suburban community of Bang Bua Thong, Nontaburi Province were sampled in late June and July. Of 1,477 children, 33/1,000 had recent dengue infection and 7/1,000 had recent JE infection. The rate of dengue infection in each village influenced the rate in schools, in that the rate of the school could be predicted from the proportion of students coming from each village. This result suggested that most transmission occurred in the residential environment; otherwise, the rate in each village going to a single school would be identical. Serum samples were taken in late August in the agricultural community of Hua Samrong, Chachoengsao Province. Of 748 students in two schools, 95/1,000 had signs of recent dengue infection and 32/1,000 had signs of recent JE infection. Two of 12 villages had significantly less flavivirus infection than some other villages and three villages had significantly more flavivirus infection. The children from one village had a dengue infection rate of 256 per 1,000, which was higher than the national average for the worst year (1987) previously recorded in Thailand. Within Hua Samrong, there was evidence for significant dengue transmission in one of the schools and concentrated transmission in small areas of two of the villages. The younger age group (3-8 years old) had significantly higher risk of infection by either flavivirus than older children. Elevated homes with wooden floors had significantly higher risk of dengue in the largest village. The observations from 1989 describe the epidemiological situation in rapidly developing, rural villages. This stage of development is probably being repeated throughout Southeast Asia as formerly isolated, rural villages become connected by transportation and economy to urban centers. What appears to be a single dengue outbreak based on passive surveillance conducted on a regional basis may actually be a variety of epidemiological situations. The practical implication of this conclusion is that application of a combination of vaccination and vector control should be targeted to higher risk areas in order to increase the likelihood of regional dengue virus eradication.